Recently, a server system as aiming at data acquisition by the connection from a client to a server via a network has been widely used. A typical one is the Internet access mode.
Recently, as these server systems, a blade system including a plurality of server blades has been used often.
Also, as aiming at handling rapid increase in the access or keeping fault tolerance, a request from the client is balanced (distributed) for each server blade for processing often. A device for the distribution processing is called a load balancer.
In the load balance, for each server blade configuring the blade system, a method of uniformly balancing the load is generally used.
For example, methods such as “the minimum access mode” which selects a server blade having the smallest number of connections from/to the client and “the fastest response mode” which selects a server blade having the shortest response time for the request are used often.
However, in this case, in spite of a processable method by a small number of server blades when performances of the server blades are high, it is required to activate all server blades regardless of their processing amounts. Therefore, there is a possibility that power consumption of entire system is increased.
Accordingly, as a new method against the general load-balancing methods, the one disclosed in, for example, Japanese Patent Application Laid-Open Publication No. 2003-281008 (Patent Document 1) is cited.
The one disclosed in Patent Document 1 distributes the request from the client preferentially to a server having the shortest operation time, and besides, can start up only the least required servers for the entire server system by comparing a data request amount from the client with a data supply possible amount.
Patent Document 1: Japanese Patent Application Laid-Open Publication No. 2003-281008